Merge remote-tracking branch 'origin/master'

Conflicts:
	TODO

.gitignore

@@ -125,4 +125,5 @@ doc/reference/sgml.stamp
 doc/reference/xml/
 doc/html
 doc/pdf
+gtkdocerrors
 

ChangeLog

@@ -1,6 +1,10 @@
 26/7/11 started 7.26.1
 - doc fixups
 - oops, ==0 missing from a strcmp() in vips7compat
+- fixed a race in im_XYZ2Lab() table build
+- added im_concurrency_get() to operation db
+- better benchmarkn.sh runs for the correct number of CPUs automatically, runs
+  three times for each one, and just reports the fastest
 
 26/7/11 started 7.26.0
 - version bunp for 7.26

TODO

@@ -8,22 +8,6 @@
 
 - in benchmark, try:
 
-	vips im_benchmarkn temp.v temp2.v 1
-	vips --vips-concurrency=99 --vips-tile-width=10 --vips-tile-height=10 \
-		im_benchmarkn temp.v temp2a.v 1
-	vips im_avg temp2a.v
-	vips im_avg temp2.v
-
-  get different results :-(
-
-  7.24 seems to fail this test too argh
-
-  first column of tiles is fine, but stuff to the right of that is broken
-
-
-
-
-- leak check, again
 
 - vips operation print could show operation flags as well, cf. "vips
   im_subtract"
@@ -102,7 +86,6 @@
 	- do we bundle "convert" in the OS X / win32 builds? if we don't we
 	  should
 
-	- some way for nip2 to get the vips bin area
 
 
 

benchmark/benchmarkn.sh

@@ -1,4 +1,4 @@
-#!/bin/sh
+#!/bin/bash
 
 uname -a
 gcc --version
@@ -7,7 +7,7 @@ vips --version
 # how large an image do you want to process? 
 # sample2.v is 290x442 pixels ... replicate this many times horizontally and 
 # vertically to get a highres image for the benchmark
-tile=2
+tile=13
 
 # how complex an operation do you want to run?
 # this sets the number of copies of the benchmark we chain together:
@@ -21,26 +21,44 @@ if [ $? != 0 ]; then
   echo "build of test image failed -- out of disc space?"
   exit 1
 fi
-echo -n "test image is" `header -f Xsize temp.v` 
+echo -n "test image is" `header -f width temp.v` 
-echo " by" `header -f Ysize temp.v` "pixels"
+echo " by" `header -f height temp.v` "pixels"
+max_cpus=`vips im_concurrency_get`
 
+echo "max cpus = $max_cpus"
 echo "starting benchmark ..."
-echo "chain=$chain"
+echo /usr/bin/time -f %e vips \
-
+  --vips-concurrency=xx \
-for cpus in 1 2 3 4 ; do
+  --vips-tile-width=64 --vips-tile-height=64 \
-  export IM_CONCURRENCY=$cpus
+  im_benchmarkn temp.v temp2.v $chain
-
+echo reported real-time is best of three runs
-  echo IM_CONCURRENCY=$IM_CONCURRENCY
+echo cpus real-time
-  echo time -p vips im_benchmarkn temp.v temp2.v $chain
-  time -p vips im_benchmarkn temp.v temp2-$cpus.v $chain
 
+for((cpus = 1; cpus <= max_cpus; cpus++)); do
+  t1=`/usr/bin/time -f %e vips \
+	  --vips-concurrency=$cpus \
+	  --vips-tile-width=64 --vips-tile-height=64 \
+	  im_benchmarkn temp.v temp2.v $chain 2>&1`
   if [ $? != 0 ]; then
     echo "benchmark failed -- install problem?"
     exit 1
   fi
+  t2=`/usr/bin/time -f %e vips \
+	  --vips-concurrency=$cpus \
+	  --vips-tile-width=64 --vips-tile-height=64 \
+	  im_benchmarkn temp.v temp2.v $chain 2>&1`
+  t3=`/usr/bin/time -f %e vips \
+	  --vips-concurrency=$cpus \
+	  --vips-tile-width=64 --vips-tile-height=64 \
+	  im_benchmarkn temp.v temp2.v $chain 2>&1`
 
-  # find pixel average ... should be the same for all IM_CONCURRENCY settings
+  # echo $t1 $t2 $t3
-  # or we have some kind of terrible bug
+
-  echo vips im_avg temp2-$cpus.v
+  if [[ $t2 < $t1 ]]; then
-  vips im_avg temp2-$cpus.v
+	  t1=$t2
+  fi
+  if [[ $t3 < $t1 ]]; then
+	  t1=$t3
+  fi
+  echo $cpus $t1
 done

libvips/colour/im_LabQ2disp.c

@@ -1,10 +1,6 @@
-/* @(#) Turn Lab 32bit packed format into displayable rgb. Fast, but very
+/* Turn Lab 32bit packed format into displayable rgb. Fast, but very
- * @(#) inaccurate: for display only!
+ * inaccurate: for display only! Note especially that this dithers and will
- * @(#) 
+ * give different results on different runs.
- * @(#) Usage:  
- * @(#)   int im_LabQ2disp( IMAGE *in, IMAGE *out, struct im_col_display *d )
- * @(#) 
- * @(#) Returns: -1 on error, else 0
  *
  * 5/11/97 Steve Perry
  *	- adapted from old ip code

libvips/colour/im_XYZ2Lab.c

@@ -15,6 +15,8 @@
  *	- ahem, build the LUT outside the eval thread
  * 2/11/09
  * 	- gtkdoc
+ * 3/8/11
+ * 	- fix a race in the table build
  */
 
 /*
@@ -74,15 +76,14 @@ imb_XYZ2Lab_tables( void )
 {
 	static int built_tables = 0;
 
-	int was_built;
 	int i;
 
 	g_mutex_lock( im__global_lock );
-	was_built = built_tables;
+
-	built_tables = 1;
+	if( built_tables ) {
-	g_mutex_unlock( im__global_lock );
+		g_mutex_unlock( im__global_lock );
-	if( was_built )
 		return;
+	}
 
 	for( i = 0; i < QUANT_ELEMENTS; i++ ) {
 		float Y = (double) i / QUANT_ELEMENTS;
@@ -92,6 +93,10 @@ imb_XYZ2Lab_tables( void )
 		else 
 			cbrt_table[i] = cbrt( Y );
 	}
+
+	built_tables = 1;
+
+	g_mutex_unlock( im__global_lock );
 }
 
 /* Process a buffer of data.

libvips/deprecated/package.c

@@ -435,6 +435,35 @@ static im_function version_desc = {
 	version_args 			/* Arg list */
 };
 
+/* im_concurrency_get() args.
+ */
+static im_arg_desc concurrency_get_args[] = {
+	IM_OUTPUT_INT( "concurrency" )
+};
+
+/* Call im_concurrency_get() via arg vector.
+ */
+static int
+concurrency_get_vec( im_object *argv )
+{
+	int *out = ((int *) argv[0]);
+
+	*out = vips_concurrency_get();
+
+	return( 0 );
+}
+
+/* Description of im_concurrency_get.
+ */ 
+static im_function concurrency_get_desc = {
+	"im_concurrency_get", 			/* Name */
+	"get concurrency level",		/* Description */
+	0,					/* Flags */
+	concurrency_get_vec, 			/* Dispatch function */
+	VIPS_NUMBER( concurrency_get_args ), 	/* Size of arg list */
+	concurrency_get_args 			/* Arg list */
+};
+
 /* im_cache() args.
  */
 static im_arg_desc cache_args[] = {
@@ -518,6 +547,7 @@ static im_function binfile_desc = {
 static im_function *iofuncs_list[] = {
 	&binfile_desc,
 	&cache_desc,
+	&concurrency_get_desc,
 	&getext_desc,
 	&guess_prefix_desc,
 	&guess_libdir_desc,